Peripheral T-Cells, B-Cells, and Monocytes from Multiple Sclerosis Patients Supplemented with High-Dose Vitamin D Show Distinct Changes in Gene Expression Profiles.

Vitamin D is a steroid hormone that has been widely studied as a potential therapy for multiple sclerosis and other inflammatory disorders. Pre-clinical studies have implicated vitamin D in the transcription of thousands of genes, but its influence may vary by cell type. A handful of clinical studies have failed to identify an in vivo gene expression signature when using bulk analysis of all peripheral immune cells. We hypothesized that vitamin D's gene signature would vary by immune cell type, requiring the analysis of distinct cell types. Multiple sclerosis patients (n = 18) were given high-dose vitamin D (10,400 IU/day) for six months as part of a prospective clinical trial (NCT01024777). We collected peripheral blood mononuclear cells from participants at baseline and again after six months of treatment. We used flow cytometry to isolate three immune cell types (CD4+ T-cells, CD19+ B-cells, CD14+ monocytes) for RNA microarray analysis and compared the expression profiles between baseline and six months. We identified distinct sets of differentially expressed genes and enriched pathways between baseline and six months for each cell type. Vitamin D's in vivo gene expression profile in the immune system likely differs by cell type. Future clinical studies should consider techniques that allow for a similar cell-type resolution.

Generalizable Sample-Efficient Siamese Autoencoder for Tinnitus Diagnosis in Listeners With Subjective Tinnitus.

Electroencephalogram (EEG)-based neurofeedback has been widely studied for tinnitus therapy in recent years. Most existing research relies on experts' cognitive prediction, and studies based on machine learning and deep learning are either data-hungry or not well generalizable to new subjects. In this paper, we propose a robust, data-efficient model for distinguishing tinnitus from the healthy state based on EEG-based tinnitus neurofeedback. We propose trend descriptor, a feature extractor with lower fineness, to reduce the effect of electrode noises on EEG signals, and a siamese encoder-decoder network boosted in a supervised manner to learn accurate alignment and to acquire high-quality transferable mappings across subjects and EEG signal channels. Our experiments show the proposed method significantly outperforms state-of-the-art algorithms when analyzing subjects' EEG neurofeedback to 90dB and 100dB sound, achieving an accuracy of 91.67%-94.44% in predicting tinnitus and control subjects in a subject-independent setting. Our ablation studies on mixed subjects and parameters show the method's stability in performance.